This
document
represents
the
best
judgment
of
the
EPA
staff
conducting
analyses
to
support
the
Phase
II
Section
316(
b)
rule
at
this
time.
It
does
not
represent
the
final
position
of
the
Administrator
or
EPA
on
this
rulemaking.
EPA
reserves
the
right
to
change
or
amend
any
of
the
statements
within
this
document
up
until
the
time
that
the
EPA
Administrator
takes
final
action
on
this
rule.

EPA
Response
to
UWAG's
July
22,
2002,
Questions
August
1,
2002
This
memo
provides
EPA's
responses
to
the
industry
questions
on
the
Delaware
Bay
Case
Study
received
on
July
22,
2002.
The
following
provides
a
short
version
of
each
industry
question
and
EPA's
response
to
each
question.
Attachment
1
presents
the
original
version
of
the
questions
that
includes
excerpts
from
the
SAS
program
used
in
the
Delaware
Bay
case
study.

Responses
to
316
(
b)
Delaware
Bay
Case
Study
Questions
Question
1:
What
is
the
explanation
for
changing
zip
codes?

Answer:
The
ZipFip
software
used
for
estimating
distances
from
respondent
hometown
zip
codes
to
recreation
sites
used
a
zip
code
database
from
1990.
The
zip
codes
19255,
21244,
19904,
and
19092
were
not
present
in
the
database.
Rather
than
drop
these
observations
due
to
inability
to
generate
distances,
the
zip
codes
were
matched
to
the
nearby
zip
codes
for
the
same
city/
town
which
were
present
in
the
ZipFip
database.
The
number
of
affected
respondent
observations
was
twenty
(
20)
for
zip
code
19904.
No
observations
matched
the
other
zip
codes.

Question
2:
Are
trips
with
round
trip
distance
greater
than
150
miles
or
300
miles
deleted?

Answer:
Observations
with
a
round
trip
distance
greater
than
300
miles
(
or
a
one­
way
distance
greater
than
150
miles)
are
deleted
from
the
analysis.

Question
3:
What
is
the
rationale
for
adding
costs
only
to
the
chosen
sites?

Answer:
No
expenses
are
added
to
any
site.

Question
4:
What
is
the
purpose
of
price1,
price1a,
price2,
and
price3?

Answer:
These
variables
were
not
used
in
the
proposal
analysis.
They
were
created
during
previous
model
runs
and
sensitivity
analyses.

Question
5:
How
are
waterbody
size
variables
used
for
ocean
trips?
Are
there
trips
other
than
saltwater
trips
being
considered?

Answer:
The
waterbody
size
variable
is
not
used
in
the
Delaware
Bay
model.
Its
calculation
is
historical
code
from
other
model
runs.
No
trips
other
than
saltwater
were
considered.
This
document
represents
the
best
judgment
of
the
EPA
staff
conducting
analyses
to
support
the
Phase
II
Section
316(
b)
rule
at
this
time.
It
does
not
represent
the
final
position
of
the
Administrator
or
EPA
on
this
rulemaking.
EPA
reserves
the
right
to
change
or
amend
any
of
the
statements
within
this
document
up
until
the
time
that
the
EPA
Administrator
takes
final
action
on
this
rule.

Question
6:
What
is
the
rationale
for
random
draws?

Answer:
No
random
draws
are
occurring
for
the
Delaware
Case
Study.

Question7:
Why
do
program
comments
seem
to
say
that
the
draw
ends
at
7?

Answer:
The
comments
were
not
updated
for
the
current
implementation
of
the
model
code.
These
comments
should
have
been
removed.

Question
8:
Why
are
distances
set
to
½
mile
rather
than
set
to
the
mean
or
calculated
correctly?

Answer:
This
code
is
also
historical
in
nature
and
is
irrelevant
to
the
Delaware
Case
study.

Question
9:
Is
there
a
relationship
between
the
following
macro
variables?
%
let
policyadd
=
0.24;
%
let
poltext
=
'
I&
E,
Bass
Unit
Increase';

Answer:
The
macro
variable
"
policyadd"
contains
the
additional
catch
rate
per
hour
associated
with
a
1
unit
increase
in
the
number
of
fish
caught
per
trip.
(
See
also
the
answer
to
question
10.)
The
"
poltext"
macro
variable
simply
holds
a
string
to
identify
on
the
output
what
policy
is
currently
being
estimated.

Question
10:
What
is
the
implication
of
0.24?

Answer:
One
of
the
policies
being
examined
is
a
unit
increase
in
the
number
of
fish
caught
per
trip.
The
catch
rate
variable
present
in
the
site
choice
model
is
the
number
of
fish
caught
per
hour.
The
appropriate
fishing
quality
change
is
to
increase
the
per
hour
catch
rate
corresponding
to
catching
an
additional
fish
per
trip.
The
average
number
of
hours
per
fishing
trip
is
4.17.
Catching
an
additional
fish
per
trip
corresponds
to
the
following
increase
in
the
catch
rate
per
hour:

1
Fish
per
Trip/
4.17
hours
=
0.24
fish
per
hour.

Question
11:
Why
are
trip
cost
and
demographic
variables
(
i.
e.,
trip
cost,
hourly
wage,
ownboat,
etc.
created
in
both
Coastal
Restricted
Choice
Sets.
sas
and
Coastal
Model
Benefits
(
bass)
unit
increase
12_
11_
01.
sas?
Are
they
created
identically?

Answer:
Yes,
these
variables
are
created
identically.
Historically,
separate
programs
generated
the
data
set
for
the
site
choice
model
and
the
benefit
model.
While
the
This
document
represents
the
best
judgment
of
the
EPA
staff
conducting
analyses
to
support
the
Phase
II
Section
316(
b)
rule
at
this
time.
It
does
not
represent
the
final
position
of
the
Administrator
or
EPA
on
this
rulemaking.
EPA
reserves
the
right
to
change
or
amend
any
of
the
statements
within
this
document
up
until
the
time
that
the
EPA
Administrator
takes
final
action
on
this
rule.

same
changes
to
the
underlying
data
set
were
made,
no
static
input
data
set
with
the
final
variables
was
created.
Thus,
the
same
transformation
must
occur
in
the
benefits
model
to
generate
the
required
variables
for
estimating
individual
welfare
gain
from
changes
in
the
fishing
quality
of
the
sites.

Question
12:
What
is
the
appropriate
interpretation
of
the
following
variables?
haus,
swelf_
b,
swelf_
p,
pctlow,
pctup,
baselow,
difflow,
basehgh,
diffhgh,
pcttrip
Answer:
The
variables
haus,
swelf_
b,
swelf_
p,
pctlow,
pctup,
baselow,
difflow,
basehgh,
and
diffhgh
are
various
measures
of
seasonal
welfare
and
changes
in
seasonal
welfare.
These
variables
are
not
used
in
the
analysis
of
the
Delaware
Bay.
Because
we
used
a
generic
benefits
program,
many
variables
and
statistics
are
generated
which
may
not
be
relevant
to
a
specific
analysis.

The
pcttrip
variable
represents
the
percentage
change
in
the
number
of
trips
taken
per
season
by
an
individual
as
a
result
of
the
policy
change.
This
value
is
calculated
as
follows:

((
estimated
number
of
trips
under
the
policy
scenario
­
estimated
number
of
baseline
trips)/
estimated
number
of
baseline
trips)*
100
Question
13:
How
does
the
output
of
benefits
estimation
programs
link
with
spreadsheets
4­
1031,
4­
1032,
and
4­
1033?

Answer:
Please
see
the
attached
spreadsheet
and
SAS
output
file.
The
benefits
estimation
programs
generate
average
per
trip
benefits
by
species
and
policy
scenario
as
well
as
the
estimated
average
percent
increase
in
the
number
of
trips
due
to
changes
in
fishing
circumstances.
These
values
(
per
trip
benefits
and
percent
increase
in
trips)
are
then
manually
entered
into
the
spreadsheets,
which
are
set
to
estimate
overall
benefits
based
on
the
per
trip
benefits
and
the
percentage
increase
in
the
number
trips.

For
example,
per
trip
benefits
for
the
weakfish
anglers
based
on
changes
in
I&
E
at
the
Salem
plant
only
are
estimated
as
follows:
°
The
percentage
increase
in
the
catch
rate
based
on
reduced
I&
E
from
Salem
is
added
to
1
and
inputted
in
the
benefits
program
4­
1014
as
the
macro
variable
"
policymult."
°
The
output
from
the
benefit
estimation
yields
a
mean
value
for
"
welf,"
which
is
the
per
trip
welfare
increase,
and
a
mean
value
for
the
percentage
increase
in
trips
"
pcttrip."
This
document
represents
the
best
judgment
of
the
EPA
staff
conducting
analyses
to
support
the
Phase
II
Section
316(
b)
rule
at
this
time.
It
does
not
represent
the
final
position
of
the
Administrator
or
EPA
on
this
rulemaking.
EPA
reserves
the
right
to
change
or
amend
any
of
the
statements
within
this
document
up
until
the
time
that
the
EPA
Administrator
takes
final
action
on
this
rule.

°
The
mean
"
welf"
value
for
weakfish
is
placed
in
cell
B18
of
the
worksheet
"
Est
Benefits
PrivateRental
Boat"
and
the
mean
percent
increase
in
trips
for
weakfish
anglers
is
placed
in
cell
E18
of
worksheet
"
Est
Benefits
PrivateRental
Boat."

Similar
benefits
runs
are
done
for
striped
bass
and
no
target
anglers
and
the
estimated
values
placed
in
the
worksheet
"
Est
Benefits
PrivateRental
Boat."
The
following
example
maps
the
spreadsheet
cell
values.

Worksheet
"
Est
Benefits
PrivateRental
Boat"
in
the
Excel
spreadsheet.

Species
Per
trip
Benefit
%
increase
in
#
trips
Weakfish
B18
E18
Striped
bass
B19
E19
No
Target
B20
E20
Per
trip
benefits
and
percent
increase
in
trips
are
assumed
to
be
equivalent
for
charter
boat
anglers
and
shore
anglers.
Thus,
the
same
cells
as
listed
above
on
sheets
"
Est
Benefits
Shore
Anglers"
and
"
Est
Benefits
Charter
Boats"
reference
back
to
the
sheet
"
Est
Benefits
PrivateRental
Boat."

Question
14a:
Which
outputs
from
SAS
benefits
programs
or
cells
from
spreadsheets
4­
1031,
4­
1032,
and
4­
1033
correspond
to
numbers
in
the
report
text?

Answer:
Please
see
the
attached
spreadsheet
and
SAS
output
file.
Table
B5­
8
lists
the
increase
in
per
trip
welfare
under
the
various
scenarios
for
each
species.
These
values
correspond
to
cells
B18,
B19,
and
B20
of
worksheet
"
Est
Benefits
PrivateRental
Boat"
from
the
appropriate
excel
spreadsheet
(
i.
e.,
Salem
only,
In­
Scope
Facilities,
and
All
Facilities).

Tables
B5­
11,
B5­
12,
and
B5­
13
contain
estimates
located
on
the
worksheet
"
Total
Benefits
All
Modes"
from
the
excel
spreadsheet
for
the
appropriate
scenario.
The
cells
referenced
on
worksheet
"
Total
Benefits
All
Modes"
are:

Species
Low
Value
High
Value
Weakfish
B34
C34
Striped
bass
B35
C35
No
Target
B36
C36
Total
rec.
use
B38
C38
This
document
represents
the
best
judgment
of
the
EPA
staff
conducting
analyses
to
support
the
Phase
II
Section
316(
b)
rule
at
this
time.
It
does
not
represent
the
final
position
of
the
Administrator
or
EPA
on
this
rulemaking.
EPA
reserves
the
right
to
change
or
amend
any
of
the
statements
within
this
document
up
until
the
time
that
the
EPA
Administrator
takes
final
action
on
this
rule.

Question
14b:
Which
numbers
are
related
and
how
are
they
calculated?

Answer:
Please
see
the
attached
spreadsheet
and
SAS
output
file.
For
a
given
policy
scenario,
the
Low
and
High
benefit
values
are
calculated
for
each
combination
of
the
fishing
mode
(
private
boat,
charter
boat,
and
shore
angler)
and
species.
Benefits
are
based
on
the
per
trip
welfare
gain
for
anglers
targeting
fish
species
of
concern,
the
baseline
number
of
fishing
trips,
and
the
estimated
percentage
increase
in
trips
due
to
changes
in
the
quality
of
fishing
sites.
The
low
benefits
value
is
based
on
the
baseline
number
of
trips
for
a
given
species
and
the
high
value
is
based
on
the
estimated
post­
compliance
number
of
trips.

The
benefits
for
a
given
species/
mode
combination
are
estimated
on
the
appropriate
worksheet
in
the
excel
spreadsheet
(
e.
g.,
"
Est
Benefits
PrivateRental
Boat,"
"
Est
Benefits
Shore
Angler").
The
estimated
benefit
value
associated
with
improved
catch
rates
for
a
given
species
is
calculated
by
multiplying
the
number
of
trips
taken
by
anglers
targeting
this
species
by
the
benefit
per
trip
from
improved
catch
rates
for
the
relevant
fish
species.

The
total
number
of
baseline
fishing
trips
is
the
average
total
number
of
trips
for
a
given
fishing
mode
in
DE
and
NJ
for
the
years
1994
to
1998
multiplied
by
the
percentage
of
trips
targeting
a
given
species
based
on
the
MRFSS
(
Marine
Recreational
Fisheries
Statistics
Survey)
data.

For
example,
Low
weakfish
benefits
for
Private
and
Rental
boaters
are
calculated
as
follows:

[(
Average
total
number
of
fishing
trips
taken
by
private/
rental
boat
anglers
per
year
DE
*
%
of
trips
by
private/
rental
boat
anglers
targeting
weakfish
DE)
+
(
Average
total
number
of
fishing
trips
taken
by
private/
rental
boat
anglers
per
year
NJ
*
%
of
trips
by
private/
rental
boat
anglers
targeting
weakfish
NJ)]*
Per
trip
welfare
gain
for
anglers
targeting
weakfish.

This
calculation
is
done
for
the
other
two
modes
­­
Charter
Boats
and
Shore
Anglers.
Benefits
for
weakfish
anglers
are
then
summed
over
the
three
fishing
modes
to
generate
the
overall
benefits
for
weakfish
anglers
(
cell
B34).
Estimations
of
the
High
benefits
value
(
cell
C34)
relies
on
the
same
equation
as
above
but
multiplies
the
total
number
of
trips
per
year
by
the
estimated
percentage
increase
in
trips
due
to
improved
weakfish
catch
rates
(
cell
E18).
This
document
represents
the
best
judgment
of
the
EPA
staff
conducting
analyses
to
support
the
Phase
II
Section
316(
b)
rule
at
this
time.
It
does
not
represent
the
final
position
of
the
Administrator
or
EPA
on
this
rulemaking.
EPA
reserves
the
right
to
change
or
amend
any
of
the
statements
within
this
document
up
until
the
time
that
the
EPA
Administrator
takes
final
action
on
this
rule.

Attachment
1:
Industry
Questions
The
link
between
benefit
calculations
and
values
listed
in
EPA's
report
is
unclear.

1.
For
example,
for
4­
1011
(
Coastal
Restricted
Choice
Sets
III.
sas):

if
d2a
=
19255
then
d2a
=
19106;
/*
Philadelphia,
PA*/
if
d2a
=
21244
then
d2a
=
21240;
/*
Baltimore,
MD*/
if
d2a
=
19904
then
d2a
=
19903;
/*
Dover,
DE*/

if
d2a
=
19092
then
d2a
=
19090;
/*
Philadelphia,
PA*/

What
is
the
explanation
for
changing
zip
codes?

2.
Programs
originally
appears
to
delete
distances
greater
than
150
miles:

k=
1;
link
chosen;
if
dst(
k)
>
150
then
delete;
chdist=
dst(
k);
end;
else
delete;
/*
no
chosen
site
known
*/

Later,
trip
distances
are
doubled
for
trip
cost
calculation
if
lost_
inc=
1
then
do;
tripcst
=
(
0.29*
ds*
2)
+
((
wage*(
ds/
40))*
2);
timecst
=
0;
end;
else
do;
tripcst
=
(
0.29*
ds*
2);
timecst
=
(
ds/
40)*
2;
end;

Are
trips
with
round
trip
distance
greater
than
150
miles
or
300
miles
deleted?

3.
Notation
to
code
reads:

/*
If
it's
the
chosen
site,
add
any
travel
related
expenses
as
stated
by
the
angler
in
the
interview.
If
it's
not
the
chosen
site,
then
we
don't
know
about
any
extra
expenses,
so
nothing
is
added.
Extra
expenses
include
parking
fees,
gas,
tolls,
ferrys
*/
This
document
represents
the
best
judgment
of
the
EPA
staff
conducting
analyses
to
support
the
Phase
II
Section
316(
b)
rule
at
this
time.
It
does
not
represent
the
final
position
of
the
Administrator
or
EPA
on
this
rulemaking.
EPA
reserves
the
right
to
change
or
amend
any
of
the
statements
within
this
document
up
until
the
time
that
the
EPA
Administrator
takes
final
action
on
this
rule.

What
is
the
rationale
for
adding
costs
only
to
the
chosen
sites?

4.
The
variables
price1
­
price3
are
created
as
follows:

price1=
0.5*
op_
cost+
tr_
cost;
price1a=
0.4*
op_
cost+
tr_
cost;
price2=.
33*
op_
cost+
tr_
cost;
price3=.
25*
op_
cost+
tr_
cost;

What
is
the
purpose
of
price1,
price1a,
price2,
and
price3?

5.
Waterbody
size
variable
is
created
in
the
following
manner:

/*
Create
log
of
water
acres
as
a
size
proxy
0.00019
feet
per
mile
1
square
mile
is
639.9974
acres
*/

acres
=
(
segl*(
pwidth*
0.00019))*
639.9974;
if
missing(
acres)
=
0
and
acres
^=
0
then
ln_
acre
=
log(
acres);
/*
natural
log
*/
else
ln_
acre
=
1;

if
segl=.
or
segl=
0
then
rlen
=
10.197;
else
rlen
=
segl;
if
pwidth=.
then
rwid=
29.068;
else
rwid
=
pwidth;

How
are
waterbody
size
variables
used
for
ocean
trips?
Are
there
trips
other
than
saltwater
trips
being
considered?

6.
Notation
to
SAS
code
reads:

There
are
23
DE
sites
and
25
NJ
sites.
PA
residents
have
a
choice
set
containing
all
48
sites.

Yet
the
random
draw
code
appears
to
draw
all
sites
(
for
example,
code
below
selects
23
of
23
DE
and
MD
sites)

/*
Draw
Delaware
sites
for
Delaware
and
Maryland
Anglers
*/

if
d1b
=
'
DE'
or
d1b
=
'
MD'
then
do;
do
k
=
2
to
23;
if
k
=
2
then
do;
do
until
(
lori=
979);
lori=
lori+
1;
link
reachit;
set
reach.
attribut5(
rename=(
coast=
rchzone)
keep=
coastal
rchstate
coast)
This
document
represents
the
best
judgment
of
the
EPA
staff
conducting
analyses
to
support
the
Phase
II
Section
316(
b)
rule
at
this
time.
It
does
not
represent
the
final
position
of
the
Administrator
or
EPA
on
this
rulemaking.
EPA
reserves
the
right
to
change
or
amend
any
of
the
statements
within
this
document
up
until
the
time
that
the
EPA
Administrator
takes
final
action
on
this
rule.

point=
lori;
/*
point
to
reach
row
in
water
data
*/
if
(
rch(
k)
^=
newrchid)
&
(
coastal=
1)
&
(
rchstate='
DE')
then
do;
c(
1)
=
c(
1)
+
1;
count(
c(
1))
=
lori;
end;
/*
if
then
*/
end;
/*
do
until
*/
end;
/*
if
k
=
7
*/
start=
1;
range=
1;
if
c(
1)
<
23
then
do;
if
k
<=
(
c(
1)+
1)
then
do;
lori=
count(
k­
1);
link
reachit;
end;
else
do;
dst(
k)=.;
rch(
k)=.;
end;
/*
else
do
*/
end;
else
link
rndset;
end;
/*
do
2
to
7
*/
end;
/*
if
de
or
md
angler
*/
link
resetit;

Given
this,
what
is
the
rationale
for
random
draw?

7.
Also,
when
it
appears
from
the
code
that
23
sites
are
drawn
(
do
k
=
2
to
23;)
why
do
program
comments
seem
to
say
that
the
draw
ends
at
7?
/*
do
2
to
7
*/

8.
In
preparing
files
for
Limdep,
it
appears
that
missing
distances
are
set
to
½
mile.

data
concat;
set
combine;
array
reach{
48}
rch1­
rch48;
array
dist{
48}
dst1­
dst48;
do
I=
1
to
48;
lk=
reach{
I};
ds=
dist{
I};
if
ds=.
then
ds=.
5;
/*
missing
distance
*/
output;
end;

Why
are
distances
set
to
½
mile
rather
than
set
to
the
mean
or
This
document
represents
the
best
judgment
of
the
EPA
staff
conducting
analyses
to
support
the
Phase
II
Section
316(
b)
rule
at
this
time.
It
does
not
represent
the
final
position
of
the
Administrator
or
EPA
on
this
rulemaking.
EPA
reserves
the
right
to
change
or
amend
any
of
the
statements
within
this
document
up
until
the
time
that
the
EPA
Administrator
takes
final
action
on
this
rule.

calculated
correctly?

9.
4_
1015
(
Coastal
Model
Benefits
Bass
Unit
Increase.
sas)

Is
there
a
relationship
between
the
following
macro
variables?

%
let
policyadd
=
0.24;
%
let
poltext
=
'
I&
E,
Bass
Unit
Increase';

10.
What
is
the
implication
of
.24?
If
it
represents
a
unit
increase
in
striped
bass
catch,
what
are
the
underlying
calculations?

11.
Why
are
trip
cost
and
demographic
variables
(
ie
trip
cost,
hourly
wage,
own
boat,
etc.
created
in
both
Coastal
Restricted
Choice
Sets.
sas
and
Coastal
Model
Benefits
(
bass)
unit
increase
12_
11_
01.
sas?
Are
they
created
identically?

12.
What
is
the
appropriate
interpretation
of
the
following
variables?

var
haus
swelf_
b
swelf_
p
pctlow
pctup
baselow
difflow
basehgh
diffhgh
pcttrip;

13.
How
does
the
output
of
benefits
estimation
programs
link
with
spreadsheets
4­
1031,
4­
1032,
and
4­
1033?

14.
Which
outputs
from
SAS
benefits
programs
or
cells
from
spreadsheets
4­
1031,4­
1032,
and
4­
1033
correspond
to
numbers
in
the
report
text?
Which
numbers
are
related
and
how
are
they
calculated?
